OBJECTIVES

A NMR-based metabolomics study of minced pork meat inoculated with Brochothrix thermosphacta,

Leuconostoc gelidum and Pseudomonas fragi

CAUCHIE E.*1, LEENDERS J.2, BARÉ G.1, TAHIRI A.1, DELHALLE L.1, KORSAK N.1, DE TULLIO P.2 AND DAUBE G.11Food Science Department, Food Microbiology, FARAH, ULiège.

2Center for Interdisciplinary Research on Medicines (CIRM), Laboratory of Medicinal Chemistry, ULiège, 4000 Liège, Belgium.

FARAH Day 2017Liège, October 13th 2017

INTRODUCTION

* ecauchie@ulg.ac.be

In order to control food waste, studies have highlighted the importance of

monitoring the microbial diversity of food products.

CONCLUSIONSExploration of the correlation of these metabolites with microbial counts (Spoilage Value (Sval) at 7.0 log CFU.g-1) suggested their use as possible spoilage indicators. These results support the use of NMR-based metabolomics as a valuable tool to provide information on pork meat spoilage and to follow intrinsically the evolution of the metabolomics pattern linked to a specific strain in a complex bacterial ecosystem.

The aim of the current study was to assess meat spoilage through theevolution of bacterial counts and changes in the metabolic profile ofminced pork meat using Proton Nuclear Magnetic Resonance (1H-NMR).

MATERIALS AND METHODS

4°C

8°C

12°C

2 packaging conditions

*daily analysis** analysis at day 0 and at day 13

Classical microbiology for Total Viable Counts (TVA) onPCA at 22°C for inoculated* (n=705) and non inoculatedsamples** (n=55)

Gamma irradiation(10 kGy)

Brochothrix thermosphactaLeuconostoc gelidumPseudomonas fragi

3 dominantbacterial strains

RESULTS AND DISCUSSION

in triplicate

Based on previous agingtests (data not shown) Inoculation by 3 log CFU.g-1 for

each bacterial strain

Analysis at day 0 and at day 13

storage

Non-inoculatedsamples

5 repetitions(n=210)

100 mg of meat

+ 500 µl D2O

Homogenize

with a vortex 1

min

Centrifugation

2x30 sec

5000 rpm

Add 300 µl D2O

Homogenize

with a vortex 1

min

Leave for 15 min

Centrifugation

15 min

13000 rpm

Analysis of

surpernatant

(650 µl) by 1H-NMR

Samples discrimination into 3 groups based on their spectral pattern is observed for each bacterial strain (1, 2, 3) and between the type of strain inoculated (4)..

Samplespreparation

RMN Bruker 500 Mhz

Spectral analysis

Statistical analysis(OPLS-DA)

Metabolomics pattern

Metabolites

1. Food packaging 2. Temperature 3. Non-inoculated VS inoculated

R2=0,735; Q2=0,685 FWMAP

Example: Brochothrix thermosphacta Example: Pseudomonas fragi

R2=0,715; Q2=0,670 Non-inoculatedInoculated

Example: Leuconostoc gelidum

4°C8°C12°C

R2=0,662; Q2=0,602 R2=0,842; Q2=0,830 B. thermosphactaP. fragiL. gelidum

B. thermosphacta P. fragi L. gelidum Non-inoculatedAcetateGlycerol

Statiscal metabolites descriptors explaining the difference between groups: somemetabolites are significatly increased (p<0,05) for each inoculated samples by bacterialstrains (bacterial metabolic activity), and for the non-inoculated samples (naturaldegradation of minced pork meat).

ThreonineGlycine

BetaineLactate

LactateLipoproteins

1H-NMRMetabolomics

GlucoseTaurine

irradiated meat

Challenge testFor inoculated and non-

inoculated samples

13 days

Proton Nuclear Magnetic resonance (1H-NMR)**

(n=210)

Mixing of samples

Kenwood Mixer speed 2 for 2 min

Food wrap (FW) Modified Atmosphere Packaging (MAP) CO2 30 % - O2 70 %

0,0

1,0

2,0

3,0

4,0

5,0

6,0

7,0

8,0

9,0

10,0

11,0

0 1 2 3 4 5 6 7 8 9 10 11 12 13

Ba

ce

ria

l co

un

ts (

log

CF

U/g

)

Time (days)

Example of B. thermosphacta in MAP

4

8

12

FRESH

SPOILED

Estimation of the Minimal Shef Life (MSL) of samples which isdefined by the time when TVC reached 7.0 log CFU.g-1 (SpoilageValue, Sval).

A Sval is obtained for each conditions and could be statisticallycorrelate to metabolomics results in further analysis (data notavailable)

MSL

Sval

Association to bacterial counts

B. thermosphactaL. gelidumP. fragi

4. Type of strain inoculated